Journal of Cardiovascular Computed Tomography 14 (2020) 400–406

Contents lists available at ScienceDirect

Journal of Cardiovascular Computed Tomography

journal homepage: www.elsevier.com/locate/jcct

Research paper Percent atheroma volume: Optimal variable to report whole-heart T atherosclerotic plaque burden with coronary CTA, the PARADIGM study Alexander R. van Rosendaela,b, Fay Y. Lina, Xiaoyue Mac, Inge J. van den Hoogena,b, Umberto Giannia,d, Omar Al Husseina, Subhi J. Al'Arefa, Jessica M. Peñaa, Daniele Andreinie, Mouaz H. Al-Mallahf, Matthew J. Budoffg, Filippo Cademartirih, Kavitha Chinnaiyani, Jung Hyun Choij, Edoardo Contee, Hugo Marquesk, Pedro de Araújo Gonçalvesk, Ilan Gottliebl, Martin Hadamitzkym, Jonathon A. Leipsicn, Erica Maffeio, Gianluca Pontonee, Gilbert L. Raffi, Sanghoon Shinp, Yong-Jin Kimq, Byoung Kwon Leer, Eun Ju Chuns, Ji Min Sungt,u, Sang-Eun Leet,u, Daniel S. Bermanv, Renu Virmaniw, Habib Samadyx, Peter H. Stoney, ∗ Jagat Narulaz, Jeroen J. Baxb, Leslee J. Shawa, James K. Mina, , Hyuk-Jae Changt a Department of Radiology, NewYork-Presbyterian Hospital and Weill Cornell Medicine, New York, NY, USA b Department of , Leiden University Medical Center, Leiden, the Netherlands c Department of Healthcare Policy and Research, New York-Presbyterian Hospital and the Weill Cornell Medical College, New York, NY, USA d Department of Molecular Medicine, Section of Cardiology, University of Pavia, Pavia, Italy e Centro Cardiologico Monzino, IRCCS Milan, Italy f Houston Methodist DeBakey Heart & Vascular Center, Houston Methodist Hospital, Houston, TX, USA g Department of Medicine, Los Angeles Biomedical Research Institute, Torrance, CA, USA h Cardiovascular Imaging Center, SDN IRCCS, Naples, Italy i Department of Cardiology, William Beaumont Hospital, Royal Oak, MI, USA j Pusan University Hospital, Busan, South Korea k UNICA, Unit of Cardiovascular Imaging, Hospital da Luz, Lisboa, Portugal l Department of Radiology, Casa de Saude São Jose, Rio de Janeiro, Brazil m Department of Radiology and Nuclear Medicine, German Heart Center Munich, Munich, Germany n Department of Medicine and Radiology, University of British Columbia, Vancouver, BC, Canada o Department of Radiology, Area Vasta 1/ASUR Marche, Urbino, Italy p Division of Cardiology, Department of Internal Medicine, Ewha Womans University Seoul Hospital, Seoul, South Korea q Department of Internal Medicine, Seoul National University College of Medicine, Cardiovascular Center, Seoul National University Hospital, Seoul, South Korea r Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea s Seoul National University Bundang Hospital, Sungnam, South Korea t Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Yonsei University Health System, Seoul, South Korea u Yonsei-Cedars-Sinai Integrative Cardiovascular Imaging Research Center, Yonsei University College of Medicine, Yonsei University Health System, South Korea v Department of Imaging and Medicine, Cedars Sinai Medical Center, Los Angeles, CA, USA w Department of , CVPath Institute, Gaithersburg, MD, USA x Division of Cardiology, Emory University School of Medicine, Atlanta, GA, USA y Cardiovascular Division, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA z Icahn School of Medicine at Mount Sinai, Mount Sinai Heart, Zena and Michael A. Wiener Cardiovascular Institute, Marie-Josée and Henry R. Kravis Center for Cardiovascular Health, New York, NY, USA

ARTICLE INFO ABSTRACT

Keywords: Background and aims: Different methodologies to report whole-heart atherosclerotic plaque on coronary com­ puted tomography (CCTA) have been utilized. We examined which of the three commonly used Imaging plaque burden definitions was least affected by differences in body surface area (BSA) and sex. Percent atheroma volume Methods: The PARADIGM study includes symptomatic patients with suspected coronary atherosclerosis who Coronary CTA underwent serial CCTA > 2 years apart. Coronary , vessel, and plaque were quantified from the coronary tree on a 0.5 mm cross-sectional basis by a core-lab, and summed to per-patient. Three quantitative methods of plaque burden were employed: (1) total plaque volume (PV) in mm3, (2) percent atheroma volume (PAV) in %

∗ Corresponding author. E-mail address: [email protected] (J.K. Min). https://doi.org/10.1016/j.jcct.2020.01.012 Received 20 August 2019; Received in revised form 5 November 2019; Accepted 29 January 2020 Available online 30 January 2020 1934-5925/ © 2020 Published by Elsevier Inc. on behalf of Society of Cardiovascular Computed Tomography A.R. van Rosendael, et al. Journal of Cardiovascular Computed Tomography 14 (2020) 400–406

3 [which equaled: PV/vessel volume * 100%], and (3) normalized total atheroma volume (TAVnorm) in mm [which equaled: PV/vessel length * mean population vessel length]. Only data from the baseline CCTA were

used. PV, PAV, and TAVnorm were compared between patients in the top quartile of BSA vs the remaining, and between sexes. Associations between vessel volume, BSA, and the three plaque burden methodologies were assessed. Results: The study population comprised 1479 patients (age 60.7 ± 9.3 years, 58.4% male) who underwent CCTA. A total of 17,649 coronary segments were evaluated with a median of 12 (IQR 11–13) segments per-patient (from a 16-segment coronary tree). Patients with a large BSA (top quartile), compared with the

remaining patients, had a larger PV and TAVnorm, but similar PAV. The relation between larger BSA and larger

absolute plaque volume (PV and TAVnorm) was mediated by the coronary vessel volume. Independent from the atherosclerotic risk (ASCVD) score, vessel volume correlated with PV (P < 0.001), and

TAVnorm (P = 0.003), but not with PAV (P = 0.201). The three plaque burden methods were equally affected by sex.

Conclusions: PAV was less affected by patient's body surface area then PV and TAVnorm and may be the preferred method to report coronary atherosclerotic burden.

1. Introduction data from the baseline CCTA were used.

Coronary computed tomography angiography (CCTA) allows for accurate evaluation of atherosclerotic plaque, and its presence, extent and severity are strongly related to future major adverse cardiovascular 2.2. CCTA imaging and analysis protocol events.1 More recently, studies have demonstrated the ability of CCTA to accurately quantify the total coronary plaque burden (whole-heart) CCTAs were acquired using ≥64 detector row CT scanners, with 2–5 and to assess changes in plaque burden on CCTA over time. image acquisition protocols in accordance with the Society of Plaque burden with (IVUS) has tradition­ Cardiovascular Computed Tomography guidelines, including the sys­ ally been reported using percent atheroma volume (PAV)—defined as tematic administration of beta-blocker and nitroglycerin before acqui­ the total plaque volume (PV)/vessel volume x 100%—or total atheroma sition.10 CCTA DICOM files from each participating site were ­ trans volume normalized to vessel length (TAVnorm)—defined as PV/vessel ferred to a Central Core Laboratory for blinded image analysis. Level III 6,7 length x mean population vessel length. Normalization is performed experienced CTA readers from the Central Core Laboratory, masked to to adjust for differences in the length of the interrogated coronary patient clinical and demographical characteristics, performed qualita­ segments. Compared with IVUS, CCTA offers the ability to quantify and tive and semi-quantitative CCTA analysis using dedicated software characterize atherosclerotic plaque from all coronary artery segments (QAngioCT Research Edition v2.1.9.1; Medis Medical Imaging Systems, simultaneously, albeit at a lower spatial resolution. This allows for Leiden, the Netherlands).5 CCTA imaging analysis methodology in­ whole-heart atherosclerotic plaque assessment by total PV without cluded routine measures of inter- and intra-observer variability, as has 2,3 normalization for vessel length or volume. To date, no study—either been described previously: intra-class correlation coefficient 0.992 for by CCTA or IVUS—has examined how PV, PAV or TAVnorm are influ­ inter-observer, and 0.996 for intra-observer plaque volume repeat­ enced by scenarios that affect vessel length and vessel volume. Previous ability.5,11 In brief, measures of coronary plaque, lumen and vessel wall histological and imaging data has shown that the size of is were assessed in all coronary segments ≥2 mm in diameter, using 8,9 dependent on body surface area (BSA) and sex. standardized width/level display settings adjusted for aortic contrast The ideal method for determining atherosclerotic plaque burden attenuation. Contours were manually adjusted where needed. Coronary would be least affected – most stable – to patient demographics. PV (no plaque was defined as any tissue ≥1 mm2 within or adjacent to the normalization of plaque), PAV (normalization for vessel volume), and lumen that can be distinguished from the surrounding pericardial 1 TAVnorm (normalization for vessel length) may vary according to BSA tissue, epicardial or the coronary artery lumen. Measurements were and sex – which are known to affect vessel size (length and volume). In performed per-segment on a 0.5 mm cross-sectional basis according to a a large, prospective CCTA study with whole-heart quantification of modified 16-segment coronary tree model.12 Lumen and vessel wall coronary plaque, lumen and vessel wall, we aimed to determine the contours were overlapping in segments without plaque. To obtain per- optimal way to report coronary atherosclerotic plaque burden. patient measures, the data for all coronary artery segments were summed. 2. Methods

2.1. Patients 2.3. Quantitative plaque burden reporting The PARADIGM (Progression of AtheRosclerotic PlAque DetermIned by Computed TomoGraphic Angiography IMaging) study is Coronary artery plaque burden was reported according to the fol­ a dynamic, multicenter, observational registry which prospectively lowing three previously described definitions in the CCTA and IVUS collects clinical, procedural, and follow-up data for MACE from patients literature: Plaque volume (PV), percent atheroma volume (PAV) and 5 undergoing clinically indicated CCTA. The study was approved by the total atheroma volume normalized for vessel length (TAVnorm), with institutional review board of all participating centers. The PARADIGM definitions as described below.2–4,6,13–17 registry enrolled 2252 patients from 13 sites in 7 countries (Brazil, Canada, Germany, Italy, Portugal, South Korea and the United States) Plaque volume (PV) = Summation of plaque from all segments between 2003 and 2015. Patients underwent a baseline and follow-up Percent atheroma volume (PAV) = (Plaque volume / Vessel volume) * CCTA at least 2 years apart, using ≥64-slice CT-scanners for the eva­ 100% luation of CAD. For the current analysis, patients with previous cor­ onary revascularization or uninterpretable CCTA for quantitative ana­ Total atheroma volume normalized (TAVnorm) = (Plaque volume) / lysis were excluded, leaving 1479 patients in the current study. Only Vessel length) * mean population vessel length

401 A.R. van Rosendael, et al. Journal of Cardiovascular Computed Tomography 14 (2020) 400–406

2.4. PV, PAV, and TAVnorm by the 3 different methods. Two-sided P-values < 0.05 were considered statistically significant. All analyses were performed with SAS (version PV is summation of plaque from all analyzable segments from the 9.4; SAS Institute Inc., Cary, NC) and R 3.3.0 (R Development Core 16-segment coronary tree. PAV and TAVnorm are similar to PV but have Team, 2016). been normalized for either the length of the vessel (TAVnorm) or the volume (PAV) respectively. 3. Results We examined how differences in vessel volume, body surface area (BSA), and sex affect PV, PAV, andnorm TAV . Previous imaging and 3.1. Patients histopathological data have demonstrated that the size (volume and 8,9 length) of arteries is larger with increasing BSA and male sex. Given In total, 1479 patients were included with a mean age of the different types of normalization (or no normalization for PV), al­ 60.7 ± 9.3 years and 58.4% was male. A median of 12 (IQR 11–13) terations in BSA and sex may affect PV, PAV, and TAVnorm. An example coronary segments were analyzed per patient (17,649 segments in of how PV, PAV and TAVnorm are affected by vessel volume, BSA, and total). The clinical indication for CCTA was evaluation of chest pain in sex is shown in Fig. 1. 83.1% (Table 1) The mean height, weight, BMI, and BSA were 167 ± 9.7 cm, 71 ± 13 kg, 25.3 ± 3.3 kg/m2, and 1.81 ± 0.21 m2, 2.5. Statistical analysis respectively. Most patients had non-obstructive CAD (62.5%) and 24.0% of the patients had more than 5 diseased coronary artery seg­ Continuous variables were expressed as mean ± standard devia­ ments (Table 1). tion (SD) or median (25th – 75th interquartile range) if non-normally distributed and categorical variables as counts (percentage). Because of 3.2. PV, PAV, and TAVnorm according to BSA, and sex their left skewed distribution, all plaque data are provided as medians with interquartile range. Normality of data distribution was assessed by Patients in the top quartile of vessel volume, compared with the review of histograms and Q-Q plots. Continuous variables were com­ remaining patients, had similar PAV, but greater PV and a trend to­ pared with the independent T-test or Mann-Whitney U test as appro­ wards greater TAVnorm: 2.0% (IQR 0.32–5.7) vs 2.4% (IQR 0.0–7.1) priate. Categorical variables were compared with the chi-square test as P = 0.445, 73.8 mm3 (IQR 9.8–197.6) vs 40.7 mm3 (IQR 0.0–118.7) 3 3 appropriate. The variability of PV, PAV, and TAVnorm was assessed P < 0.001, and 49.2 mm (IQR 7.2–143.4) vs 42.9 mm (IQR according to vessel volume, BSA, and sex, with dichotomizing vessel 0.0–135.1) P = 0.075, respectively (Table 2). Similarly, when cate­ volume and BSA according to their 75th and 50th percentile. PV, PAV, gorizing patients into the top quartile of BSA vs the remaining patients, and TAVnorm were compared between the BSA and sexes. The associa­ PAV was similar, but PV and TAVnorm were significantly larger in pa­ tion between BSA and vessel volume was examined with a scatter plot tients with the largest BSA, Table 3. The results remained unchanged and the Pearson's correlation coefficient was determined. The athero­ when vessel volume and BSA were dichotomized by the 50th percentile sclerotic cardiovascular disease score (ASCVD) was calculated.18 Linear (Appendix Tables 1 and 2). regression analyses were performed between vessel volume and the 3 PV, PAV, and TAVnorm were similarly affected by sex: plaque burden plaque burden methodologies (log-transformed) – adjusting for ASCVD by all 3 methodologies was higher in men, Table 4. In patients at low score – to evaluate the effect of vessel size with plaque burden defined risk (ASCVD score < 7.5) plaque burden by all 3 methodologies was

Fig. 1. An example of how vessel volume and BSA influence PV, PAV, TAVnorm. The left panel shows an RCA with 2 lesions (red coded coronary plaques) in a patient with a high BSA and vessel volume. The right panel shows an RCA with similar plaque volume (430 mm3) but smaller vessel volume. As a result of the normalization for vessel volume, PAV is larger in the RCA in the right panel, while PV and TAVnorm are similar. BSA, body surface area; PAV, percent atheroma volume; PV, plaque volume; RCA, right coronary artery; R-PL, right postero-lateral branch; R-PDA, right posterior descending artery; TAVnorm, total atheroma volume normalized for vessel length. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

402 A.R. van Rosendael, et al. Journal of Cardiovascular Computed Tomography 14 (2020) 400–406

Table 1 Table 3 Baseline characteristics. Effect of body surface area on plaque burden measurements.

Clinical Value CCTA Value (N = 1479) Above 75th percentile below 75th percentile P-value (N = 1479) (N = 369) (N = 1110)

Demographics Extent of CAD CCTA Age, years 60.7 ± 9.3 No CAD, n 334 (22.6) PV, mm3 65.2 (13.8–149.2) 41.2 (0.0–133.3) 0.001 Male, n 864 (58.4) Non-obstructive 924 (62.5) PAV, % 2.5 (0.53–6.5) 2.1 (0.0–7.0) 0.225 3 CAD, n TAVnorm, mm 59.3 (12.9–148.6) 41.0 (0.0–134.2) 0.005 Height, cm 167 ± 9.7 Obstructive CAD, 221 (14.9) Vessel volume, 2781 ± 1135 2111 + 932 < 0.001 n mm3 Weight, kg 71.3 ± 13.2 Vessel length, mm 419 ± 132 388 ± 120 < 0.001 Body mass index, 25.3 ± 3.3 Number of diseased segments BSA, m2 2.06 (2.00–2.16) 1.74 (1.62–1.84) < 0.001 kg/m2 Male sex, n 324 (88.8) 528 (48.4) < 0.001 Body surface area, 1.81 ± 0.21 0, n 334 (22.6) m2 BSA, body surface area; CCTA, coronary computed tomography angiography;

1-5, n 790 (53.4) PAV, percent atheroma volume; PV, plaque volume; TAVnorm, total atheroma Ethnicity > 5, n 335 (24.0) volume normalized for vessel length. Caucasian, n 402 (27.2) East-Asian, n 1013 (68.5) Plaque quantification Other, n 64 (4.3) PV, mm3 47.1 (3.9–138.7) Table 4 PAV, % 2.3 (0.13–6.8) Effect of sex on plaque burden measurements. 3 Cardiac symptoms TAVnorm, mm 44.0 (2.6–137.6) Male (N = 864) Female (N = 615) P-value No chest pain, n 251 (16.9) Vessel volume, 2282 ± 1024 mm3 CCTA Non-anginal, n 136 (9.2) Vessel length, 396 ± 123 PV, mm3 61.4 (10.-155.1) 29.6 (0.0–102.5) < 0.001 mm PAV, % 2.8 (0.44–7.4) 1.5 (0.0–5.7) < 0.001 Atypical, n 1021 (69.0) Lumen volume, 2174 ± 991 3 TAVnorm, mm 58.2 (9.2–154.4) 28.3 (0.0–110.4) < 0.001 mm3 Vessel volume, mm3 2428 ± 1066 2075 ± 922 < 0.001 Typical, n 71 (4.8) Vessel length, mm 404 ± 123 384 ± 123 < 0.001 Cardiovascular risk factors BSA, m2 1.90 ± 0.19 1.69 ± 0.17 < 0.001 Diabetes, n 303 (20.6) Hypertension, n 790 (53.8) BSA, body surface area; CCTA, coronary computed tomography angiography;

Dyslipidemia, n 576 (39.2) PAV, percent atheroma volume; PV, plaque volume; TAVnorm, total atheroma Familial history for 432 (29.2) volume normalized for vessel length. CAD, n Currently smoking, n 263 (17.9) similar between sexes, Appendix Table 3. Medications Aspirin, n 559 (38.4) ACE-I/ARB, n 420 (29.0) 3.3. BSA and vessel volume Beta blocker, n 421 (29.0) Calcium channel 325 (22.5) A larger BSA was significantly associated with a larger vessel vo­ blocker, n 2 , n 591 (42.2) lume (Fig. 2): Pearson's R = 0.53 P < 0.001. Every 0.1 m increase in BSA related to an approximately 250 mm3 larger vessel volume. Fur­ ASCVD score, % 9.8 (4.9–18.7) ther, a larger vessel volume was associated with increasing PV 2 2 Data provided as mean (SD), median (25th, 75th interquartile range), or counts (P < 0.001, R = 0.121) and TAVnorm (P = 0.003, R = 0.111) after (%). adjustment for ASCVD score, but no independent association between ACE-I, angiotensin converter enzyme inhibitor; ARB, angiotensin receptor vessel volume and PAV was observed (P = 0.201, R2 = 0.116), ap­ blocker; CAD, ; PAV, percent atheroma volume; PV, pendix Table 4. plaque volume; TAVnorm, total atheroma volume normalized for vessel length. 3.4. Discussion Table 2 Effect of vessel volume on plaque burden measurements. In the current study, we aimed to determine which plaque burden Vessel volume > 75th Vessel volume < 75th P-value methodology is most stable – shows the least amount of variability - percentile (N = 369) percentile (N = 1110) according to differences in BSA and sex. PV and TAVnorm were both CCTA larger in patients with the highest BSA, while PAV was not affected by 3 PV, mm 73.8 (9.8–197.6) 40.7 (0.0–118.7) < 0.001 BSA. This relationship between larger BSA and larger PV and TAVnorm PAV, % 2.0 (0.32–5.7) 2.4 (0.0–7.1) 0.445 was mediated by coronary vessel volume, which showed a parallel in­ 3 TAVnorm, mm 49.2 (7.2–143.4) 42.9 (0.0–135.1) 0.075 crease with larger BSA (Pearsons R = 0.52, P < 0.001). Vessel volume, 3660 ± 744 1823 ± 609 < 0.001 mm3

Vessel length, 509 ± 88 358 ± 109 < 0.001 3.5. PV, PAV, and TAVnorm to measure atherosclerotic burden mm BSA, m2 1.93 ± 0.19 1.77 ± 2.0 < 0.001 Male sex, n 258 (69.9) 606 (54.6) < 0.001 The clinical expert consensus documentation on reporting plaque 19 burden with IVUS recommends to use either PAV or TAVnorm. CCTA BSA, body surface area; CCTA, coronary computed tomography angiography; differs compared to IVUS by the ability to quantify plaque fromall PAV, percent atheroma volume; PV, plaque volume; TAVnorm, total atheroma coronary segments which resulted in the use of PV: summation of volume normalized for vessel length. whole-heart atherosclerotic plaque without normalization.2,4 PV, PAV,

and TAVnorm will differ between each other when the length orthe volume of are not exactly similar (Fig. 1), stressing the need for standardization in reporting. Previous imaging and histological studies showed that BSA and sex affect vessel size (function of vessel

403 A.R. van Rosendael, et al. Journal of Cardiovascular Computed Tomography 14 (2020) 400–406

Fig. 2. Scatter plots between per patient vessel volume and BSA. Among patients without coronary plaque (N = 359), increasing BSA was associated with increasing vessel volume (Pearson's correlation coefficients 0.53, P < 0.001). The equation of the linear fitting line was −2360 + 2520*BSA. As such, a 0.12 m increase in BSA relates to a 252 m3 higher vessel volume. Restricted was to patients without coronary plaque to avoid the enlarging effect of plaque on vessel volume (positive remodeling).

length and area). We studied how PV, PAV, and TAVnorm are affected by that PAV is superior to report coronary plaque burden. BSA and sex, while the optimal way is most stable – least variable – Although men had a significantly larger vessel volume than women, according to differences in BSA and sex. no differences were observed between the 3 plaque burden methodol­ ogies when comparing men vs women. This suggests that PV, PAV, and TAV are equally influenced by the sex of a patient. 3.6. PV, PAV, and TAVnorm according to vessel volume, BSA, and sex norm

We observed that increasing vessel volume relates to larger absolute 3.7. Limitations plaque volume as measured by PV and TAVnorm. Patients in the top quartile of vessel volume had greater PV and TAVnorm compared with Ideally, histologically measured plaque burden would serve as gold the remaining patients, while PAV was similar because of its normal­ standard to determine the accuracy of overall plaque burden mea­ ization. Given the correlation between vessel volume and BSA (corre­ surements with PV, PAV and TAVnorm. Vessel volume increases with 24 lation coefficient: 0.53), similar results were observed when PV, PAV, growth of coronary plaque (positive remodeling). Hence, when di­ and TAVnorm were compared between patients in the top quartile of BSA viding plaque volume by vessel volume as done in PAV, the true plaque and the remaining. These findings indicate that plaque volume in­ burden will be underestimated. Future studies should evaluate and creases proportionally with the size of the vessel, and indirectly with define the true volume of diseased coronary arteries as if there was no BSA. BSA has been previously associated with aortic dimensions9 or left plaque (i.e. not affected by positive remodeling) which can be used to atrium,20 and these structures are usually adjusted for BSA to reduce normalize the quantified plaque volume for. Most patients were low variability in the measurement and define ‘normal’ values. Pathological risk and had East-Asian ethnicity which might limit generalizability of data in explanted hearts demonstrated a similar relationship between a the findings. No comparisons of prognostic value between the plaque larger BSA and increasing coronary vessel size.8 It seems therefore burden methodologies were performed. Finally, differences in vasodi­ apparent that coronary atherosclerosis demonstrates a similar positive latory response to pre-CT use of nitroglycerine use may have affected correlation with the size of the vessel, or indirectly BSA. In coronary vessel volumes. arteries, a similar volume of plaque results in greater relative reductions in coronary lumen in smaller coronary arteries compared to larger ar­ 4. Conclusion teries. Similarly, a larger volume of plaque is needed to provide a si­ milar relative reduction in lumen in larger vessels compared with We observed that PAV was less affected by patient's body surface smaller vessels. PAV is a function of vessel volume and may therefore area then PV and TAVnorm. PAV may be the preferred method to report better reflect the clinical importance of coronary atherosclerosis than coronary atherosclerotic burden in CCTA. absolute plaque quantification definitions (PV and TAVnorm). In line with this, Kishi et al. demonstrated that PAV showed the highest ac­ Funding curacy to identify hemodynamically significant compared with 21 several other plaque measurements including TAVnorm. Further, low This work was supported by the Leading Foreign Research Institute variability is important for standardization in reporting of plaque Recruitment Program through the National Research Foundation (NRF) burden and endpoint selection in the design of trials to reach statistical of Korea funded by the Ministry of Science and ICT (MSIT) (Grant No. significance. In addition, multiple IVUS trials have observed lower 2012027176). The study was also funded in part by a generous gift 6,22,23 variability of PAV compared to TAVnorm, supporting our findings from the Dalio Institute of Cardiovascular Imaging (New York, NY) and

404 A.R. van Rosendael, et al. Journal of Cardiovascular Computed Tomography 14 (2020) 400–406 the Michael Wolk Foundation (New York, NY). scientific advisory board of Arineta and GE Healthcare, and hasan equity interest in Cleerly. Dr. Habib Samady serves on the scientific Declaration of competing interest advisory board of Philips, has equity interest in Covanos Inc., and has a research grant from Medtronic. Dr. Kavitha Chinnaiyan is a non-com­ Dr. James K. Min receives funding from the Dalio Foundation, pensated medical advisory board member of Heartflow Inc. The­ re National Institutes of Health, and GE Healthcare. Dr. Min serves on the maining authors have no relevant disclosures.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jcct.2020.01.012.

APPENDIX

Table 1 Effect of vessel volume on plaque burden measurements

Vessel volume > 50th percentile (N = 740) Vessel volume < 50th percentile (N = 739) P-value

CCTA PV, mm3 64.1 (7.1–173.0) 33.2 (0–105.2) < 0.001 PAV, % 2.2 (0.22–6.3) 2.3 (0–7.4) 0.504 3 TAVnorm, mm 47.7 (5.2–139.7) 39.3 (0–133.1) 0.060 Vessel volume, mm3 3067 ± 806 1496 ± 460 < 0.001 Vessel length, mm 473 ± 90 319 ± 103 < 0.001 BSA, m2 1.89 ± 0.19 1.74 ± 0.20 < 0.001 Male sex, n 479 (64.7) 385 (52.1) < 0.001

BSA, body surface area; CCTA, coronary computed tomography angiography; PAV, percent atheroma volume; PV, plaque volume; TAVnorm, total atheroma volume normalized for vessel length.

Table 2 Effect of body surface area on plaque burden measurements

BSA > 50th percentile (N = 728) BSA < 50th percentile (N = 727) P-value

CCTA PV, mm3 55.8 (7.5–150.1) 40.6 (0–124.6) 0.004 PAV, % 2.3 (0.34–6.6) 2.3 (0–7.1) 0.652 3 TAVnorm, mm 47.2 (6.7–141.9) 39.4 (0–128.9) 0.040 Vessel volume, mm3 2634 ± 1097 1924 ± 812.2 < 0.001 Vessel length, mm 419 ± 129 373 ± 113 < 0.001 BSA, m2 1.98 ± 0.13 1.65 ± 0.11 < 0.001 Male sex, n 583 (80.1) 269 (37.0) < 0.001

BSA, body surface area; CCTA, coronary computed tomography angiography; PAV, percent atheroma volume; PV, plaque volume; TAVnorm, total atheroma volume normalized for vessel length.

Table 3 Effect of sex on plaque burden measurements in low risk patients (ASCVD score < 7.5%)

Male (N = 256) Female (N = 328) P-value

CCTA PV, mm3 25.4 (0.0–117.2) 17.6 (0.0–72.8) 0.145 PAV, % 1.3 (0.0–4.7) 0.93 (0.0–3.7) 0.256 3 TAVnorm, mm 24.1 (0.0–97.2) 15.8 (0.0–63.1) 0.198 Vessel volume, mm3 2497 ± 1026 2139 ± 967 < 0.001 Vessel length, mm 429 ± 110 403 ± 119 0.009 BSA, m2 1.91 ± 0.18 1.70 ± 0.16 < 0.001

ASCVD, atherosclerotic cardiovascular disease; BSA, body surface area; CCTA, coronary computed tomography angiography; PAV, percent

atheroma volume; PV, plaque volume; TAVnorm, total atheroma volume normalized for vessel length.

405 A.R. van Rosendael, et al. Journal of Cardiovascular Computed Tomography 14 (2020) 400–406

Table 4 Linear association between vessel volume and plaque burden

Standardized beta-coefficient for P-value Standardized beta-coefficient for P-value Standardized beta-coefficient for P-value PV* TAVnorm* PAV*

ASCVD score, % 0.307 < 0.001 0.327 < 0.001 0.338 < 0.001 Vessel volume, mm3 0.173 < 0.001 0.074 0.003 −0.031 0.201

*Log-transformed. ASCVD, atherosclerotic cardiovascular disease; BSA, body surface area; CCTA, coronary computed tomography angiography; PAV, percent atheroma volume; PV, plaque volume; TAVnorm, total atheroma volume normalized for vessel length.

References Tomogr. 2009;3:122–136. 13. Papadopoulou SL, Neefjes LA, Garcia-Garcia HM, et al. Natural history of coronary atherosclerosis by multislice computed tomography. JACC Cardiovasc Imag. 1. Min JK, Shaw LJ, Devereux RB, et al. Prognostic value of multidetector coronary 2012;5:S28–S37. computed tomographic angiography for prediction of all-cause mortality. J Am Coll 14. Ceponiene I, Nakanishi R, Osawa K, et al. Coronary artery calcium progression is Cardiol. 2007;50:1161–1170. associated with coronary plaque volume progression: results from a quantitative 2. Hauser TH, Salastekar N, Schaefer EJ, et al. Effect of targeting with semiautomated coronary artery plaque analysis. JACC Cardiovasc Imag. 2018 salsalate: the TINSAL-CVD randomized clinical trial on progression of coronary Dec;11(12):1785–1794. plaque in overweight and obese patients using . JAMA Cardiol. 15. Nakanishi R, Ceponiene I, Osawa K, et al. Plaque progression assessed by a novel 2016;1:413–423. semi-automated quantitative plaque software on coronary computed tomography 3. Budoff MJ, Ellenberg SS, Lewis CE, et al. Testosterone treatment and coronary artery angiography between diabetes and non-diabetes patients: a propensity-score plaque volume in older men with low testosterone. Jama. 2017;317:708–716. matching study. Atherosclerosis. 2016;255:73–79. 4. Shin S, Park HB, Chang HJ, et al. Impact of intensive LDL lowering on 16. Nissen SE, Nicholls SJ, Sipahi I, et al. Effect of very high-intensity statin therapy on coronary artery atherosclerosis progression: a serial CT angiography study. JACC regression of coronary atherosclerosis: the ASTEROID trial. Jama. Cardiovasc Imag. 2017;10:437–446. 2006;295:1556–1565. 5. Lee SE, Chang HJ, Sung JM, et al. Effects of statins on coronary atherosclerotic 17. Gu H, Gao Y, Hou Z, et al. Prognostic value of coronary atherosclerosis progression plaques: the PARADIGM (progression of AtheRosclerotic PlAque DetermIned by evaluated by coronary CT angiography in patients with stable . Eur Radiol. computed TomoGraphic angiography imaging) study. JACC Cardiovasc Imag. 2018 2018;28:1066–1076. Oct;11(10):1475–1484. 18. Goff Jr DC, Lloyd-Jones DM, Bennett G, et al. 2013 ACC/AHA guideline on the as­ 6. Nissen SE, Tuzcu EM, Libby P, et al. Effect of antihypertensive agents on cardio­ sessment of cardiovascular risk: a report of the American college of cardiology/ vascular events in patients with coronary disease and normal : the American heart association task force on practice guidelines. J Am Coll Cardiol. CAMELOT study: a randomized controlled trial. Jama. 2004;292:2217–2225. 2014;63:2935–2959. 7. Nissen SE, Nicholls SJ, Wolski K, et al. Effect of rimonabant on progression of 19. Mintz GS, Garcia-Garcia HM, Nicholls SJ, et al. Clinical expert consensus document atherosclerosis in patients with abdominal obesity and coronary artery disease: the on standards for acquisition, measurement and reporting of intravascular ultrasound STRADIVARIUS randomized controlled trial. Jama. 2008;299:1547–1560. regression/progression studies. EuroIntervention : J EuroPCR Collab Work Group 8. Litovsky SH, Farb A, Burke AP, et al. Effect of age, race, body surface area, heart Interventional Cardiol Eur Soc Cardiol. 2011;6:1123–1130 9. weight and atherosclerosis on coronary artery dimensions in young males. 20. Pritchett AM, Jacobsen SJ, Mahoney DW, Rodeheffer RJ, Bailey KR, Redfield MM. Atherosclerosis. 1996;123:243–250. Left atrial volume as an index of left atrial size: a population-based study. J Am Coll 9. Kalsch H, Lehmann N, Mohlenkamp S, et al. Body-surface adjusted aortic reference Cardiol. 2003;41:1036–1043. diameters for improved identification of patients with thoracic aortic : 21. Kishi S, Magalhaes TA, Cerci RJ, et al. Total coronary atherosclerotic plaque burden results from the population-based Heinz Nixdorf Recall study. Int J Cardiol. assessment by CT angiography for detecting obstructive coronary artery disease as­ 2013;163:72–78. sociated with myocardial perfusion abnormalities. J Cardiovasc Comput Tomogr. 10. Abbara S, Blanke P, Maroules CD, et al. SCCT guidelines for the performance and 2016;10:121–127. acquisition of coronary computed tomographic angiography: a report of the society 22. Nissen SE, Tsunoda T, Tuzcu EM, et al. Effect of recombinant ApoA-I Milano on of cardiovascular computed tomography guidelines committee: endorsed by the coronary atherosclerosis in patients with acute coronary syndromes: a randomized north American society for cardiovascular imaging (NASCI). J Cardiovasc Comput controlled trial. Jama. 2003;290:2292–2300. Tomogr. 2016;10:435–449. 23. Nissen SE, Tuzcu EM, Schoenhagen P, et al. Effect of intensive compared with 11. Lee SE, Chang HJ, Rizvi A, et al. Rationale and design of the Progression of moderate -lowering therapy on progression of coronary atherosclerosis: a ran­ AtheRosclerotic PlAque DetermIned by Computed TomoGraphic Angiography domized controlled trial. Jama. 2004;291:1071–1080. IMaging (PARADIGM) registry: a comprehensive exploration of plaque progression 24. Sipahi I, Tuzcu EM, Schoenhagen P, et al. Compensatory enlargement of human and its impact on clinical outcomes from a multicenter serial coronary computed coronary arteries during progression of atherosclerosis is unrelated to atheroma tomographic angiography study. Am Heart J. 2016;182:72–79. burden: serial intravascular ultrasound observations from the REVERSAL trial. Eur 12. Raff GL, Abidov A, Achenbach S, et al. SCCT guidelines for the interpretation and Heart J. 2006;27:1664–1670. reporting of coronary computed tomographic angiography. J Cardiovasc Comput

406